Synthesis of amphidinolide y precursors

Article abstract of DOI:10.1016/j.tetlet.2013.12.047

The Negishi coupling between a chiral C3 synthon and an iodoalkene arising from 3-butyn-1-ol, which gave the C3-C9 fragment of amphidinolide Y, was the starting point of a formal total synthesis of this marine natural product. By means of Sharpless ADH and TADDOL-mediated crotylation, the full western fragment (C1-C11) was obtained, which was coupled with the eastern fragment (3-hydroxyoxolane derivative). The penultimate step (ring-closing metathesis, with G-II, H-G-II, or Nitro-Grela reagents, under several conditions) posed great difficulties. The cyclization was achieved with 15c (7,9-bis-O-TES) and 15d (7-O-TES, 9-O-TBS); more than stoichiometric amounts of the H-G-II Ru complex were required for complete conversion.

Full text of DOI:10.1016/j.tetlet.2013.12.047

Tetrahedron Letters 55 (2014) 900–902
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Synthesis of amphidinolide Y precursors
⇑
⇑
Laura Mola, Anna Olivella, Fèlix Urpí , Jaume Vilarrasa
Departament de Química Orgànica, Facultat de Química, Universitat de Barcelona, Diagonal 645, 08028 Barcelona, Catalonia, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
The Negishi coupling between a chiral C3 synthon and an iodoalkene arising from 3-butyn-1-ol, which
gave the C3–C9 fragment of amphidinolide Y, was the starting point of a formal total synthesis of this
marine natural product. By means of Sharpless ADH and TADDOL-mediated crotylation, the full western
fragment (C1–C11) was obtained, which was coupled with the eastern fragment (3-hydroxyoxolane
derivative). The penultimate step (ring-closing metathesis, with G-II, H–G-II, or Nitro-Grela reagents,
under several conditions) posed great difficulties. The cyclization was achieved with 15c (7,9-bis-O-
TES) and 15d (7-O-TES, 9-O-TBS); more than stoichiometric amounts of the H–G-II Ru complex were
required for complete conversion.
Received 7 November 2013
Revised 7 December 2013
Accepted 10 December 2013
Available online 19 December 2013
Keywords:
Formal total synthesis
Amphidinolides
Ring-closing metathesis (RCM)
Macrocyclizations
Ó 2013 Elsevier Ltd. All rights reserved.
Amphidinolide Y (1) is a cytotoxic macrolide that was isolated
from a marine Amphidinium dinoflagellate by Kobayashi et al.1
Two total syntheses have been reported:2 the first based on the for-
mation of the C12–C13 bond by a variant of Suzuki coupling and
Yamaguchi macrolactonization,2a while the second relies upon a
ring-closing metathesis (RCM) approach to create the C11–C12
double bond.2b We describe here a third synthesis, also with a
RCM reaction as the penultimate step. When we designed total
syntheses of 1 and its biogenetic derivative1 amphidinolide X3
we faced such a disconnection (Scheme 1). The stereo selective for-
mation of trisubstituted double bonds by RCM was (and is) chal-
lenging, but the shortcomings of the methods are a stimulus to
improve them. Compound 1 might behave as a G-actin assembly
inhibitor, as amphidinolides X and J.4
The eastern fragment (C12–C21, Scheme 1) was not expected to
be a problem, since we had prepared several closely related syn-
thons and precursors of the terminal olefin (Y = CHO, CN, COMe,
C„CH, CH@CH2) and stereoisomers, when working on the synthe-
sis of amphidinolide X.3a,b Problems were expected to appear in the
coupling of the two fragments. The failure to achieve the desired
(E)-double bond in the RCM leading to the 16-membered ring of
amphidinolide X had forced us to opt for a Si-tethered CM reaction
that, unfortunately, involved several further steps,3a,5 followed by a
final macrolactonization. As 1 has a slightly larger size (17-mem-
bered ring), the chances of success might be slightly higher either
by a direct RCM,6 through cascade or relay RCM (RRCM),6c,d or via
other types of RCM.7
In fact, the publication of Dai et al.2b (40% of the desired RCM
product) when our project was starting reinforced our initial strat-
egy of adopting the shortest approach (direct RCM). We report here
one successful synthesis of the western fragment (Scheme 1), its
union with the eastern fragment by esterification, and the efforts
to carry out the final RCM. Most attempts were unfruitful, but in
our opinion they deserve to be reported as an evaluation of the
scope of the current RCM methods when applied to densely func-
tionalized substrates.
First of all, we synthesized the C3–C9 enantiopure fragment (4)
shown in Scheme 2, starting from commercially available 3-butyn-
1-ol and methyl (R)-3-hydroxy-2-methylpropanoate (Roche’s
ester, 99% ee). The key step, the formation of bond C5–C6 between
2 and 3 by a Negishi reaction, according to careful conditions for
the preparation of the organozinc,8 took place in high yield.
The asymmetric dihydroxylation (Sharpless’ ADH, see Scheme 3)
of 4 with AD-mix-b9 was complete, but diol 5 was contaminated
with 20% of its diastereomer (dihydroxylation by the opposite
face). Separation was performed by flash chromatography. We
observed later that it was easier after conversion of the mixture
into the isopropylidene acetals (6 and its stereomer).
RCM
11
11
12
O
16
9
12
HO
HO
HO
HO
21
PGO
13
O
+
or
O
16
O
OH
Y
O
1
5
O
16
O
O
3
21
PGO
eastern fragment
amphidinolide Y (1)
western fragment
⇑
Corresponding authors. Tel.: +34 934021258; fax: +34 933397878 (J.V.).
E-mail address: jvilarrasa@ub.edu (J. Vilarrasa).
Scheme 1. Main fragments of amphidinolide Y.
0040-4039/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetlet.2013.12.047

L. Mola et al. / Tetrahedron Letters 55 (2014) 900–902
901
Removal of PMB10 from 6 with DDQ, followed by Swern oxida-
tion11 of alcohol 7 and a standard Wittig reaction, gave fragment
C1–C9 (8).12 Removal of the TBDPS group to give 912 was followed
by another Swern oxidation. To secure the anti-relationship of the
hydroxy group at C9 and methyl at C10 (see 10) we applied the
crotylation procedure of Hafner–Duthaler et al.13 Other known
asymmetric allylation and crotylation procedures, which we have
used in other syntheses, could have been checked and compared,
but the method with Ti-TADDOL-ate13 was sufficiently efficient
in the first trials, with anti/syn ratios between 90:10 and 95:5.
Stereoisomer 10 was purified by column chromatography.
We protected the free hydroxy group of 10 as its TBS ether, but
later attempts to hydrolyze the 1,3-dioxolane (isopropylidene ace-
tal) without removing the TBS group were unsuccessful. The oxida-
tion of the secondary OH group at C6 should not be postponed until
the end of the synthesis, as we had noted a tendency of the free C6-
OH to add to the double bond (conjugate addition under base catal-
ysis) to afford a stable oxolane derivative. It therefore seemed wise
to oxidize C6-OH to ketone as soon as possible, while C9-OH was
protected. Thus, as shown in Scheme 4, the hydroxy group at C9
was protected with TIPS (11), which allowed the selective cleavage
of the isopropylidene acetal14 to yield 12. The Swern oxidation,
conversion of the ester group of 13 into carboxylic acid 14 (under
special conditions,15 to prevent the conjugate addition of the
C7-OH to the double bond with formation of a THP/oxane ring),
activation of the COOH group with ethoxyethyne and [RuCl2(p-
cymene)]2,16 and reaction with the appropriate eastern fragment3
(with the hydroxy group free) in the presence of 10-camphorsulfo-
nic acid gave 15a. A portion of 15a was deprotected with TBAF/
AcOH in THF to afford 15b. Protection of a sample of 15b with an
excess of TESOTf and 2,6-lutidine gave 15c.
AlMe3
Cp2ZrCl2
I2
OTBDPS
92%
OH
OTBDPS
CH2Cl2
rt, 15 h
THF
I
2
0
2 h
O
OH
I
LitBu
ZnCl2
Pd(PPh3)4
THF, rt, 15 h
MeO
OPMB
THF, rt
60%
20 min,
rt, 1 h
3
OTBDPS
9
5
3
OPMB
85%
4
Scheme 2. Synthesis of fragment C3–C9.
OTBDPS
DDQ
OTBDPS MeO OMe
O
HO
AD-mix-
O
4
OH
CH3SO2NH2
CH2Cl2
pH 8 buffer
TsOH
CH2Cl2, rt
10 min
t
OPMB
OPMB
2O
48 h
0
99% (4:1)
5
6
OTBDPS
Swern
OTBDPS
OMe
O
O
9
Ph3P=CHCOOMe
CH2Cl2, rt, 48 h
TBAF, AcOH
O
O
6
2
THF,
rt, 3 h
OH
99%
O
7
8
98% (2 steps)
dr 99:1
Ph Ph
O
O
O
We were ready to examine the RCM reactions of 15a–c with
three available initiators (Fig. 1).6 A long series of trials, each with
5 mg of substrate, were performed by addition of up to 60 mol % of
these initiators in three batches, either in CH2Cl2 at 40 °C for 3 days
or in toluene at 90 °C for 2 days, with or without p-benzoquinone17
as an additive. The screening was better followed by electrospray
ionization MS (ESIMS).18
With the TIPS derivative (15a) no reaction was noted. With
unprotected 15b no reaction occurred either.19 The most encourag-
ing result was with 15c and H–G-II (up to 60 mol %) in toluene at
Ti
O
Cl
OH
Ph
Ph
O
O
O
11
MgCl
Swern
O
OH
Et2O,
6 h
OMe
1
OMe
O
O
9
100%
10
77% (2 steps)
95:5
Scheme 3. From fragment C3–C9 to C1–C11.
9
9
CH2Cl2
(100:4:4)
2O
TIPSOTf
2,6-lutidine
HO
HO
O
O
OTIPS
OTIPS
6
6
10
CH2Cl2
rt, 5 h
4.5 h
OMe
OMe
3
O
O
93%
12
97%
11
Swern
HO
OEt
11
Ru(II)
toluene
rt, 2 h
HO
HO
AlBr3
THT
OTIPS
O
OTIPS
OTIPS
O
O
O
OH
OMe
eastern
fragment
CSA, CH2Cl2
rt, 2 h
O
O
O
O
56%
11
71%
15a
87%
14
13
11
11
9
TESO
TESO
HO
6
12
OH
OTBS
O
OTES
O
O
O
O
O
O
O
O
3
O
O
O
15d
15b
15c
Scheme 4. Esters 15a–d.

902
L. Mola et al. / Tetrahedron Letters 55 (2014) 900–902
11
Supplementary data
12
9
PG'-O
PG-O
MesN
Cl
NMes MesN
Cl
NMes
MesN
Cl
NMes
O
Supplementary data (1H and 13C NMR spectra of the new
compounds) associated with this article can be found, in the online
version, at http://dx.doi.org/10.1016/j.tetlet.2013.12.047.
7
Ru
O
Ru
Ru
O
Cl
Cl
O
Cl
O
Ph
NO2
PCy3
iPr
O
iPr
Nitro-Grela
Grubbs-II
(G-II)
16c, PG = PG' = TES
References and notes
16d, PG = TES, PG' = TBS
1. Tsuda, M.; Izui, N.; Shimbo, K.; Sato, M.; Fukushi, E.; Kawabata, J.; Kobayashi, J.
J. Org. Chem. 2003, 68, 9109–9112.
Figure 1. Initiators/catalysts examined. Expected products 16c and 16d.
2. (a) Fürstner, A.; Kattnig, E.; Lepage, O. J. Am. Chem. Soc. 2006, 128, 9194–9204;
(b) Jin, J.; Chen, Y.; Li, Y.; Wu, J.; Dai, W.-M. Org. Lett. 2007, 9, 2585–2588; for a
review, see: (c) Fürstner, A. Isr. J. Chem. 2011, 51, 329–345; for the
bioevaluation of analogues of amphidinolides Y and X, see: (d) Fürstner, A.;
Kattnig, E.; Kelter, G.; Fiebig, H.-H. Chem. Eur. J. 2009, 15, 4030–4043.
3. (a) Rodríguez-Escrich, C.; Urpí, F.; Vilarrasa, J. Org. Lett. 2008, 10, 5191–5194;
(b) Rodríguez-Escrich, C.; Olivella, A.; Urpí, F.; Vilarrasa, J. Org. Lett. 2007, 9,
989–992; (c) Rodríguez-Escrich, C. PhD Thesis; Universitat de Barcelona, 2008;
(d) Olivella, A. PhD Thesis; Universitat de Barcelona, 2012.
90 °C for 2 days, but the signal that may be attributable to 16c
(ESIMS, m/z 696.51, M+NHþ4 , contaminated or not with its Z isomer)
was less intense than that of remaining 15c (m/z 724.54, M+NHþ4 ).
We had to subject this mixture to a second round for a complete
conversion (overall P1.2 equiv of the H–G-II reagent), which made
the isolation and purification of the product very difficult.
For comparison purposes, we prepared a few mg of the known
precursor 15d (by reaction of 15b with TBSOTf/2,6-lutidine and
then with TESOTf/2,6-lutidine). As mentioned in the introduction,
Dai et al.2b subjected 15d to a RCM reaction, using 50 mol % of
G-II in refluxing CH2Cl2 for 3 days, and isolated 16d (C11–C12 dou-
ble bond of E configuration) in 40% yield (but 0% with the Schrock
catalyst). By deprotection of the silyl ethers, these authors obtained
1. Thus, having 15d in our hands we had accomplished a formal to-
tal synthesis of 1. Nevertheless, we repeated the experiment with
5 mg of 15d and 60 mol % of G-II (added in three portions) in
refluxing CH2Cl2 for 3 days. The ESIMS peaks of the desired com-
pound (16d, m/z 696.51, M+NHþ4 ) and of 15d (m/z 724.53,
M+NHþ4 ) were of similar intensity. Thus, under our conditions, part
of 15d remained unreacted.
4. Trigili, C.; Pera, B.; Barbazanges, M.; Cossy, J.; Meyer, C.; Pineda, O.; Rodríguez-
Escrich, C.; Urpí, F.; Vilarrasa, J.; Díaz, J. F.; Barasoain, I. ChemBioChem 2011, 12,
1027–1030.
5. For reviews of Si-tethered RCM reactions, see: (a) Cusak, A. Chem. Eur. J. 2012,
18, 5800–5824; (b) Evans, P. A. In Metathesis in Natural Product Synthesis; Cossy,
J., Arseniyadis, S., Meyer, C., Eds.; Wiley–VCH: Weinheim, 2010; pp 225–259.
6. For very recent general reviews, see: (a) Deraedt, C.; d’Halluin, M.; Astruc, D.
Eur. J. Inorg. Chem. 2013, 4881–4908; (b) Olszewski, T. K.; Bieniek, M.;
Skowerski, K.; Grela, K. Synlett 2013, 24, 903–919; (c) Porta, M.; Blechert, S.
In Metathesis in Natural Product Synthesis; Cossy, J., Arseniyadis, S., Meyer, C.,
Eds.; Wiley–VCH: Weinheim, 2010; pp 313–341 (cascade CM); (d) Hoye, T. R.;
Jeon, J. In Metathesis in Natural Product Synthesis; Cossy, J., Arseniyadis, S.,
Meyer, C., Eds.; Wiley–VCH: Weinheim, 2010; pp 261–285 (relay RCM); (e)
Gradillas, A.; Pérez-Castells, J. In Metathesis in Natural Product Synthesis; Cossy,
J., Arseniyadis, S., Meyer, C., Eds.; Wiley–VCH: Weinheim, 2010; pp 149–182
(natural products); (f) Prunet, J. Eur. J. Org. Chem. 2011, 3634–3647 (natural
products).
7. For example, via ring-closing alkyne metatheses (RCAM). For reviews, see: (a)
Fürstner, A. Chem. Commun. 2011, 47, 6505–6511; (b) Wu, X.; Tamm, M.
Beilstein J. Org. Chem. 2011, 7, 82–93; (c) Davies, P. W. In Metathesis in Natural
Product Synthesis; Cossy, J., Arseniyadis, S., Meyer, C., Eds.; Wiley–VCH:
Weinheim, 2010; pp 205–223.
8. (a) Lai, K. W.; Paquette, L. A. Org. Lett. 2008, 10, 2115–2118; (b) Smith, A. B.;
Beauchamp, T. J.; LaMarche, M. J.; Kaufman, M. D.; Qiu, Y.; Arimoto, H.; Jones, D.
R.; Kobayashi, K. J. Am. Chem. Soc. 2000, 122, 8654–8664.
9. AD-mix-b: Sharpless, K. B.; Amberg, W.; Bennani, Y. L.; Crispino, G. A.; Hartung,
J.; Jeong, K.-S.; Kwong, H.-L.; Morikawa, K.; Wang, Z.-M.; Xu, D.; Zhang, X.-L. J.
Org. Chem. 1992, 57, 2768–2771. In our case (from 4 to 5) Super-AD-mix-b
afforded the same outcome.
10. Review: Wuts, P. G. M.; Greene, T. W. Protective Groups in Organic Synthesis 4th
Ed.; Wiley: Hoboken, 2007. 123–130.
11. Mancuso, A. J.; Huang, S.-L.; Swern, D. J. Org. Chem. 1978, 43, 2480–2482.
12. In a previous attempt to obtain 8 and 9, we first prepared their 2,6-diene
derivatives and treated them with AD-mix-b and Super-AD-mix-b, but the
regioselectivity was poor in all cases (around 60:40 at best).
13. (a) Hafner, A.; Duthaler, R. O.; Marti, R.; Rihs, G.; Rothe-Streit, P.;
Schwarzenbach, F. J. Am. Chem. Soc. 1992, 114, 2321–2336; for a review on
Seebach’s TADDOL uses, see: (b) Pellissier, H. Tetrahedron 2008, 64, 10279–
10317; for recent reviews of asymmetric allylations and crotylations, see: (c)
Fatima, A.; Robello, L. G.; Pilli, R. A. Quim. Nova 2006, 29, 1009–1026; (d)
Leighton, J. L. Aldrichim. Acta 2010, 43, 3–12.
We performed a final experiment with the remaining amount of
15d but with up to ca. 150 mol % of H–G-II (as always added in 3
portions, in CH2Cl2 at 40 °C for 3 days). To our delight, the ESIMS
peak at m/z 724.54 disappeared completely while that at m/z
696.51 was clearly observed (HRESIMS, calcd for C38H74NO6Siþ,
2
M+NHþ4 , m/z 696.5049, found 696.5040). As in the case of 15c, an
excess of H–G-II allowed us the complete consumption of 15d.
Unfortunately, our efforts to isolate 16d2b in a pure condition were
unsuccessful, due to the decomposition products of the reagent.
In summary, we have accomplished a new synthesis of open
precursors (15a–d) of 1, which is a formal total synthesis of 1 via
15d and relies upon the elaboration of western fragment 10. Our
approach to 15b–d consists of fourteen independent steps from
4. The reluctant RCM reaction was only feasible with fully pro-
tected precursors (15c and 15d). However, more than stoichiome-
tric amounts of the H–G-II reagent were required for a complete
conversion, which is very inconvenient. It is urgent to develop no-
vel RCM reagents capable of forming trisubstituted double bonds
embedded in natural macrolides and other complex macrocycles
under high-dilution conditions, with 610 mol % of catalyst, as well
as, although it was not the problem in this particular case, capable
of producing a stereoselective cyclization.
14. Smith, A. B.; Doughty, V. A.; Sfouggatakis, C.; Bennett, C. S.; Koyanagi, J.;
Takeuchi, M. Org. Lett. 2002, 4, 783–786.
15. Node, M.; Nishide, K.; Sai, M.; Fuji, K.; Fujita, E. J. Org. Chem. 1981, 46, 1991–
1993.
16. (a) Ohba, Y.; Takatsuji, M.; Nakahara, K.; Fujioka, H.; Kita, Y. Chem. Eur. J. 2009,
15, 3526–3537; (b) Kita, Y.; Maeda, H.; Omori, K.; Okuno, T.; Tamura, Y. J. Chem.
Soc., Perkin Trans. 1 1993, 2999–3005. and references therein; (c) Trost, B. M.;
O’Boyle, B. M. J. Am. Chem. Soc. 2008, 130, 16190–16192; (d) Trost, B. M.;
Chisholm, J. D. Org. Lett. 2002, 4, 3743–3745.
Acknowledgments
17. To prevent undesired double bond isomerizations. See: Hong, S. H.;
Sanders, D. P.; Lee, C. W.; Grubbs, R. H. J. Am. Chem. Soc. 2005, 127,
17160–17161.
Grants CTQ2006-15393, CTQ2009-13590 and 2009SGR825 are
acknowledged. A.O. enjoyed a studentship (via Fundació Bosch
Gimpera/UB, 2006–09) and was later instructor (Ajudant) in our
Department for two years, while L.M. is an UB doctorate student.
Thanks are also due to Dr Irene Fernández and Laura Ortiz (Servei
d’Espectrometria de Masses, Facultat de Química, CCiT-UB).
18. The reactions were also monitored by TLC and 1H NMR, but, due to the small
amounts of starting materials and the large amounts of initiators required,
they were better analyzed by ESIMS.
19. There seems that this cyclization is too sensitive to the conformational
preferences of the open precursors, owing to steric effects and H-bonding (and
perhaps, in the case of 15b, to the partial formation of the internal hemiketal).